From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
	id 62FC73858D38; Fri, 12 Apr 2024 09:24:57 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 62FC73858D38
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1712913897;
	bh=SrTVluxsFe6SWHbus5ZeXrvjuktjBGWjSBIAPMcaiOs=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=wka7HfZD0cwnVd9s6Pg8I6H05MAbrkMthEqZQ5+n5JOdzuQsDHQNSA7H2xlRLSs29
	 2MW+a1OICUtIi0MGEJLoUuTcrojGJUs6euuMXedgN/LbcowNvADCj7Mefdiaom2zLA
	 fjk03QUGI1a0FQ3rF9rBGRGiaHS46+d8D5NE4B44=
From: "vin.piquet at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/114695] Template argument deduction and defaulted template
 template parameters
Date: Fri, 12 Apr 2024 09:24:55 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: c++
X-Bugzilla-Version: 13.2.0
X-Bugzilla-Keywords: rejects-valid
X-Bugzilla-Severity: normal
X-Bugzilla-Who: vin.piquet at gmail dot com
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-114695-4-Y5A4r3kyn3@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-114695-4@http.gcc.gnu.org/bugzilla/>
References: <bug-114695-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
List-Id: <gcc-bugs.sourceware.org>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D114695

--- Comment #3 from Vincent Piquet <vin.piquet at gmail dot com> ---
Interesting. Now I think the issue may actually be caused by pack expansion=
 on
the call site. The issue also happens when Foo only has one base class, alb=
eit
with a different error that at least has concise types on both sides.

=3D=3D=3D=3D test.cpp =3D=3D=3D=3D
template <typename X> struct Bar {
    void accept(X value) { }
};

template <typename...> struct Foo;

template <typename... Ts>
struct Foo<Bar<Ts>...> : Bar<Ts>... {
    template <typename T, template <typename> typename B =3D Bar>
    constexpr static B<T>* slice_by_type(B<T>* self) {
        return self;
    }

    template <typename... Us>
    void accept(Us... values) {
        (slice_by_type<Us>(this)->accept(values), ...);
    }
};

int main() {
    Foo<Bar<float>> foo;
    foo.accept(3.14f);
}
=3D=3D=3D=3D test.cpp =3D=3D=3D=3D

=3D=3D=3D=3D GCC 13.2 =3D=3D=3D=3D
test.cpp: In instantiation of 'void Foo<Bar<Ts>...>::accept(Us ...) [with U=
s =3D
{float}; Ts =3D {float}]':
test.cpp:22:15:   required from here
test.cpp:16:27: error: no matching function for call to 'Foo<Bar<float>
>::slice_by_type<float>(Foo<Bar<float> >*)'
   16 |         (slice_by_type<Us>(this)->accept(values), ...);
      |          ~~~~~~~~~~~~~~~~~^~~~~~
test.cpp:10:28: note: candidate: 'template<class T, template<class> class B>
static constexpr B<T>* Foo<Bar<Ts>...>::slice_by_type(B<T>*) [with B =3D T;=
 Ts =3D
{float}]'
   10 |     constexpr static B<T>* slice_by_type(B<T>* self) {
      |                            ^~~~~~~~~~~~~
test.cpp:10:28: note:   template argument deduction/substitution failed:
test.cpp:16:27: note:   deduced conflicting types for parameter 'T' ('float'
and 'Bar<float>')
   16 |         (slice_by_type<Us>(this)->accept(values), ...);
      |          ~~~~~~~~~~~~~~~~~^~~~~~
=3D=3D=3D=3D GCC 13.2 =3D=3D=3D=3D

This may however be simplified so much that it becomes another bug. I'm uns=
ure.=